Wikidata talk:WikiProject Informatics/Software/Operating systems
Working state
editI'm searching a good property to use for the "Working state" of an OS. It would be pretty useful when searching for OSes and to avoid to include "dead" distributions.--FabC (talk) 16:09, 2 December 2017 (UTC)
The following query uses these:
- Items: operating system (Q9135)
- Properties: instance of (P31) , subclass of (P279) , discontinued date (P2669)
SELECT ?itemLabel { ?item wdt:P31/wdt:P279* wd:Q9135; wdt:P2669 []. SERVICE wikibase:label { bd:serviceParam wikibase:language "[AUTO_LANGUAGE],en". } }
There is only OS currently. But discontinued date (P2669) can be used for this purpose. John Samuel 16:37, 2 December 2017 (UTC)
- Thanks John, but unfortunately many OS distro are simply "abandoned" or more generically there are not announcements of project closures at a specific date. Maybe discontinued date (P2669) can be used as qualifier of the auspicated "Working state" property. --FabC (talk) 10:58, 3 December 2017 (UTC)
- I like the states definition as in DistroWatch:
- Active
- Dormant
- Discontinued
Comments ? --FabC (talk) 07:51, 7 December 2017 (UTC)
- operating system (Q9135), as any other system (living system (Q6506039), etc.) has state: 1) presence (Q24255051) (after creation (Q11398090)) -> 2) activity (Q1914636) -> 3) absence (Q19829125), former entity (Q15893266) (extinct language (Q38058796), extinct species (Q237350), cadaver (Q48422), etc.). So, for example, SabineOS (Q29858629) is homonym (Q160843): 1) "SabineOS (Q29858629) (-1999)" start object, 2) "SabineOS (Q29858629) (1999-2001)" working object, 3) "SabineOS (Q29858629) (2001+)" dead object. But Windows XP (Q11248) is a dead object? --Fractaler (talk) 13:56, 7 December 2017 (UTC)
- @Fractaler: In the scope of the OS life cycles, yes, Windows XP (Q11248) is a dead object since it is not supported anymore. This support the concept that the Operating System should have to be defined as subclasses and not as instances. Windows XP (Q11248) is dead despite there are still many running instances. --FabC (talk) 22:22, 9 December 2017 (UTC)
- @FabC: Yes, I can understand your problem. How about specifying novalue to discontinued date (P2669), when this information is unavailable? John Samuel 18:58, 8 December 2017 (UTC)
- @Jsamwrites: I like the idea but how can we include the element with novalue assigned to discontinued date (P2669) to the query result ? I set Ark Linux (Q1330799) accordingly but it doesn't show up --FabC (talk) 22:10, 9 December 2017 (UTC)
- @FabC: Check this
- @Jsamwrites: I like the idea but how can we include the element with novalue assigned to discontinued date (P2669) to the query result ? I set Ark Linux (Q1330799) accordingly but it doesn't show up --FabC (talk) 22:10, 9 December 2017 (UTC)
- @FabC: Yes, I can understand your problem. How about specifying novalue to discontinued date (P2669), when this information is unavailable? John Samuel 18:58, 8 December 2017 (UTC)
The following query uses these:
- Items: Linux distribution (Q131669)
- Properties: instance of (P31) , subclass of (P279) , discontinued date (P2669)
SELECT ?item WHERE { ?item wdt:P31/wdt:P279* wd:Q131669. ?item p:P2669 ?statement. ?statement rdf:type wdno:P2669. }
John Samuel 09:46, 10 December 2017 (UTC)
- Perfect! Thanks --FabC (talk) 10:29, 10 December 2017 (UTC)
Windows version hierarchy
editWhat would it be the best way to hierarchically organize the Windows OS elements and its versions ? I'm finding different and mixed approaches, each using different properties:
- Windows 8 (Q5046) is a subclass of (P279) of Windows 8.x (Q16800444), but Windows 8.x (Q16800444) is not (yet) a subclass of (P279) of Microsoft Windows (Q1406)
- Windows 8 (Q5046) is an has edition or translation (P747) of Microsoft Windows (Q1406)
- Windows 10 (Q18168774) is only a subclass of (P279) of Windows NT (Q486487) and not of Microsoft Windows (Q1406) as I would espect
- Windows 8 (Q5046) is part of the part of the series (P179) of Windows NT (Q486487)
(IMHO has edition or translation (P747) is not appropriate, but I don't want to remove something unless obviously wrong). As I'm understanding from past replies, different approaches can coexist, but each one should have a coherence. I would support the use of subclass of (P279) with Windows NT (Q486487) as root for the modern Windows OSes. I would appreciate your opinion . Thanks! --FabC (talk) 17:43, 2 January 2018 (UTC)
User interface
editSince TOPS-10 (Q1670222) showed an alert about command-line interface (Q189053) not being a graphical user interface, I replaced the GUI toolkit or framework (P1414) statement with an input device (P479) one instead. Was that perhaps premature? Looking at the en:TOPS-10 article, it now says "Default user interface" and I'm not entirely sure I'm doing it correctly. Anyway, I suggest either replacing GUI toolkit or framework (P1414) with or simply adding input device (P479) or any similar property in the infobox template to deal also with command-line operating systems (primarily historic ones developed before the 1990's). I see that the French label explicitely refers to a "graphical" interface, so maybe the different translations should be reviewed before doing this change.
Is there perhaps a WikiProject dedicated to computer history issues?--SM5POR (talk) 19:39, 22 May 2020 (UTC)